summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* [clang-tidy] Reformatted docs + minor updatesAlexander Kornienko2016-02-082-175/+223
| | | | llvm-svn: 260065
* Disable llvm/test/tools/llvm-profdata/value-prof.proftext on win32 for now. ↵NAKAMURA Takumi2016-02-071-0/+2
| | | | | | Investigating. llvm-svn: 260064
* [X86][SSE] Resolve target shuffle inputs to sentinels to permit more combinesSimon Pilgrim2016-02-073-48/+113
| | | | | | | | | | | | The combineX86ShufflesRecursively only supports unary shuffles, but was missing the opportunity to combine binary shuffles with a zero / undef second input. This patch resolves target shuffle inputs, converting the shuffle mask elements to SM_SentinelUndef/SM_SentinelZero where possible. It then resolves the updated mask to check if we have created a faux unary shuffle. Additionally, we now attempt to recursively call combineX86ShufflesRecursively for all input operands (we used to just recurse for unary integer shuffles and unary unpacks) - it safely returns early if its not a target shuffle. Differential Revision: http://reviews.llvm.org/D16683 llvm-svn: 260063
* [MS ABI] Don't emit RTTI descriptors for dllimport vtablesDavid Majnemer2016-02-072-1/+5
| | | | | | | | A dllimport'd vtable always points one past the RTTI data, this means that the initializer will never end up referencing the data. Our emission is a harmless waste. llvm-svn: 260062
* [X86][SSE] Regenerate PSHUFB shuffle mask comments testsSimon Pilgrim2016-02-071-11/+32
| | | | llvm-svn: 260061
* clang-format: [JS] Don't count shortened object literals as blocks.Daniel Jasper2016-02-072-1/+6
| | | | | | | | | | | | | | | Before: f({a}, () => { g(); // }); After: f({a}, () => { g(); // }); llvm-svn: 260060
* Revert r259961, r259978, r259981.Nico Weber2016-02-074-301/+1
| | | | | | | | | | | | The "sanitizer-windows" buildbot has been failing for two days because of this: FAILED: cl.exe asan_report.cc asan_scariness_score.h(60) : error C2536: '__asan::ScarinessScore::__asan::ScarinessScore::descr' : cannot specify explicit initializer for arrays asan_scariness_score.h(60) : see declaration of '__asan::ScarinessScore::descr' llvm-svn: 260059
* Make nozlibcompress.c pass and reenable it.Nico Weber2016-02-071-5/+1
| | | | llvm-svn: 260058
* clang-format: [JS] Support @see annotations in JSDoc comments in GoogleDaniel Jasper2016-02-071-1/+1
| | | | | | style. llvm-svn: 260057
* Disable failing nozlibcompress.cNico Weber2016-02-071-0/+4
| | | | | | | This test hasn't been running after it was added until r259976 made "REQUIRES: nozlib" work, and now that the test runs it fails. llvm-svn: 260056
* Make check line consistentDaniel Berlin2016-02-071-0/+29
| | | | llvm-svn: 260055
* [ELF][MIPS] Perform two checks using the single llvm-objdump / FileCheck runSimon Atanasyan2016-02-071-8/+8
| | | | | | NFC llvm-svn: 260054
* Revert 259942, r259943, r259948.Nico Weber2016-02-073-11/+1
| | | | | | | | | | | | | | | | The Windows bots have been failing for the last two days, with: FAILED: C:\PROGRA~2\MICROS~1.0\VC\bin\amd64\cl.exe -c LLVMContextImpl.cpp D:\buildslave\clang-x64-ninja-win7\llvm\lib\IR\LLVMContextImpl.cpp(137) : error C2248: 'llvm::TrailingObjects<llvm::AttributeSetImpl, llvm::IndexAttrPair>::operator delete' : cannot access private member declared in class 'llvm::AttributeSetImpl' TrailingObjects.h(298) : see declaration of 'llvm::TrailingObjects<llvm::AttributeSetImpl, llvm::IndexAttrPair>::operator delete' AttributeImpl.h(213) : see declaration of 'llvm::AttributeSetImpl' llvm-svn: 260053
* Test update : tighten up checksXinliang David Li2016-02-072-6/+4
| | | | llvm-svn: 260052
* Revert 260050 -- new test case should not be includedXinliang David Li2016-02-073-46/+6
| | | | llvm-svn: 260051
* Test update : tighten up checksXinliang David Li2016-02-073-6/+46
| | | | llvm-svn: 260050
* [docs] Add a note that the Visual Studio C++ tools are requiredKeno Fischer2016-02-071-0/+4
| | | | | | | | | | Watching new contributors trying to build LLVM on Windows, one of the very common failure modes was getting a version of Visual Studio that did not have a C++ compiler for CMake to put up. Trying to create a C++ project in Visual Studio will cause Visual Studio to go and download the C++ tools. llvm-svn: 260049
* [Frontend] Make the memory management of FrontendAction pointers explicit by ↵Argyrios Kyrtzidis2016-02-079-65/+73
| | | | | | using unique_ptr. llvm-svn: 260048
* [libclang] Add missing CINDEX_LINKAGE from a function.Argyrios Kyrtzidis2016-02-071-2/+2
| | | | llvm-svn: 260047
* Fix build on FreeBSD after r259741.Dimitry Andric2016-02-071-0/+4
| | | | | | | | | On FreeBSD, the uc_mcontext member of ucontext_t has a member called mc_err, which corresponds to the Linux member gregs[REG_ERR]. Reviewed by: rdivacky@FreeBSD.org llvm-svn: 260046
* Separate invariant equivalence classes by typeJohannes Doerfert2016-02-077-56/+47
| | | | | | | | | | | | | We now distinguish invariant loads to the same memory location if they have different types. This will cause us to pre-load an invariant location once for each type that is used to access it. However, we can thereby avoid invalid casting, especially if an array is accessed though different typed/sized invariant loads. This basically reverts the changes in r260023 but keeps the test cases. llvm-svn: 260045
* Make -fno-math-builtin a cc1 optionMatthew Simpson2016-02-072-2/+2
| | | | | | | | | This patch makes -fno-math-builtin a frontend only option instead of a driver option. The appropriate test case was committed in r186899 when the flag was introduced. This should fix PR26317. Contributed-by: Frank Herrmann <fgh@4gh.tv> llvm-svn: 260044
* [analyzer] Invalidate destination of std::copy() and std::copy_backward().Devin Coughlin2016-02-078-37/+241
| | | | | | | | | Now that the libcpp implementations of these methods has a branch that doesn't call memmove(), the analyzer needs to invalidate the destination for these methods explicitly. rdar://problem/23575656 llvm-svn: 260043
* [docs] Remove now confusing references to cofigure/autoconfPhilip Reames2016-02-071-10/+1
| | | | llvm-svn: 260042
* Add coverage tests (defaulted constructors/destructor)Xinliang David Li2016-02-072-0/+59
| | | | llvm-svn: 260041
* [docs] Wordsmithing to program layout descriptio in GettingStartedPhilip Reames2016-02-071-97/+63
| | | | | | | | This just incrementally improves what was already there; it's questionable whether this content belongs in the getting started guide at all. Patch by Ben Nathanson w/permission w/minor edtis by me. llvm-svn: 260040
* [docs] Clarify disk space usage of debug buildsPhilip Reames2016-02-071-8/+9
| | | | llvm-svn: 260039
* Fix a typo.Roman Divacky2016-02-071-1/+1
| | | | llvm-svn: 260038
* [docs] Remove a stale and confusing section from GettingStartedPhilip Reames2016-02-071-19/+2
| | | | | | | | The mentioned environment variable doesn't appear to have any use in the LLVM repository. If it is still relevant for clang, we can consider adding it to the clang getting started page. Patch inspired by documentation work by Ben Nathanson at the LLVM Bloomberg sprint. llvm-svn: 260037
* Revert "Re-apply r259977 - [OpenMP] Reorganize code to allow specialized ↵Renato Golin2016-02-077-92/+3
| | | | | | | | code generation for different devices." This reverts commit r259985, as it still fails one buildbot. llvm-svn: 260036
* [docs] Update the docs to describe how to build the docs with cmakePhilip Reames2016-02-072-14/+18
| | | | llvm-svn: 260035
* [X86][SSE] Added support for MOVHPD/MOVLPD + MOVHPS/MOVLPS shuffle decoding.Simon Pilgrim2016-02-0710-40/+114
| | | | llvm-svn: 260034
* [X86][AVX512] add intrinsics of Scalar FP to integer conversion with ↵Asaf Badouh2016-02-078-50/+236
| | | | | | | | rounding mode Differential Revision: http://reviews.llvm.org/D16629 llvm-svn: 260033
* [X86][SSE] Pulled out repeated target shuffle decodes into helper functions. ↵Simon Pilgrim2016-02-071-136/+89
| | | | | | | | | | NFCI. Pulled out the code used by PSHUFB/VPERMV/VPERMV3 shuffle mask decoding into common helper functions. The helper functions handle masks coming from BROADCAST/BUILD_VECTOR and ConstantPool nodes respectively. llvm-svn: 260032
* [FIX] Two "off-by-one" error in constant range usageJohannes Doerfert2016-02-0714-20/+179
| | | | llvm-svn: 260031
* Simplify code [NFC]Johannes Doerfert2016-02-072-5/+6
| | | | llvm-svn: 260030
* Use CodeGenModule::addReplacement() instead of directly accessing ↵Yaron Keren2016-02-071-1/+1
| | | | | | | | Replacements[]. This helps when trying to debug who inserted into Replacements. llvm-svn: 260028
* [ELF][MIPS] Add comments to describe __gnu_local_gp magic symbolSimon Atanasyan2016-02-072-4/+8
| | | | | | NFC. Follow-up to r259781. llvm-svn: 260027
* Fix typo in default getNoPreservedMask implementationJeroen Ketema2016-02-071-1/+1
| | | | llvm-svn: 260026
* Make memory accesses with different element types optionalTobias Grosser2016-02-0713-10/+69
| | | | | | | | We also disable this feature by default, as there are still some issues in combination with invariant load hoisting that slipped through my initial testing. llvm-svn: 260025
* AVX512: VPBROADCASTB/W/D/Q from GPR intrinsics implementation.Igor Breger2016-02-0711-134/+505
| | | | | | Differential Revision: http://reviews.llvm.org/D16813 llvm-svn: 260024
* Do not yet consider loads with non-canonical element size for load hoisting.Tobias Grosser2016-02-073-0/+47
| | | | | | | | | | Invariant load hoisting of memory accesses with non-canonical element types lacks support for equivalence classes that contain elements of different width/size. This support should be added, but to get our buildbots back to green, we disable load hoisting for memory accesses with non-canonical element size for now. llvm-svn: 260023
* Fix test case problem(caused by clang-formatXinliang David Li2016-02-072-12/+4
| | | | llvm-svn: 260022
* [PGO] add profile/coverage test cases for defaulted ctor/ctorsXinliang David Li2016-02-072-0/+88
| | | | llvm-svn: 260021
* Fix typo in comment. NFCCraig Topper2016-02-071-1/+1
| | | | llvm-svn: 260020
* Driver: adjust linker invocation for GNUToolsSaleem Abdulrasool2016-02-072-7/+42
| | | | | | | | | | | | | | | Adjust the driver to invoke the linker more similar to gcc. -dynamic-linker is only passed if -static and -shared are not part of the compiler (driver) invocation. Replicate the passing of -export-rdynamic as per the GCC link spec: %{!static: %{rdynamic:-export-dynamic} %{!shared:-dynamic-linker ...}} This behaviour is consistent across all the targets that are supported, so no need to conditionalise it on the target. Resolves PR24245. llvm-svn: 260019
* LangRef: Fix example code for cmpxchgDuncan P. N. Exon Smith2016-02-071-2/+2
| | | | | | Patch by Daniel Robertson! llvm-svn: 260018
* Sema: handle typo correction on ARC'ed ivarSaleem Abdulrasool2016-02-072-0/+12
| | | | | | | | | | | The ivar ref would be transformed by the Typo Correction TreeTransform, but not be owned, resulting in the source location being invalid. This would eventually lead to an assertion in findCapturingExpr. Prevent this assertion from triggering. Resolves PR25113. llvm-svn: 260017
* Sema: handle typo correction with ARC'ed objc propertiesSaleem Abdulrasool2016-02-072-0/+28
| | | | | | | | | | | | | We would previously assert in findCapturingExpr when performing a typo correction resulting in an assignment of an ObjC property with a strong lifetype specifier due to the expression not being rooted in the file (invalid SLoc) during the retain cycle check on the typo-corrected expression. Handle the expression type appropriately during the TreeTransform to ensure that we have a source location associated with the expression. Fixes PR26486. llvm-svn: 260016
* Don't use module context here. It's unnecessary and makes it harder to write ↵Daniel Berlin2016-02-071-2/+2
| | | | | | unittests llvm-svn: 260015
OpenPOWER on IntegriCloud